A5100323773- Industrial Vision Systems and Defect Detection
- Advanced Image and Video Retrieval Techniques
- Complex Network Analysis Techniques
- Stability and Control of Uncertain Systems
- Image Retrieval and Classification Techniques
- Advanced Sensor and Control Systems
- Adaptive Control of Nonlinear Systems
- Advanced Computational Techniques and Applications
- Opinion Dynamics and Social Influence
- Structural Health Monitoring Techniques
- Structural Load-Bearing Analysis
- Advanced Algorithms and Applications
- Advanced Neural Network Applications
- Multimodal Machine Learning Applications
- Simulation and Modeling Applications
- Human Mobility and Location-Based Analysis
- Primate Behavior and Ecology
- Chaos control and synchronization
- Concrete and Cement Materials Research
- Video Surveillance and Tracking Methods
- Blockchain Technology Applications and Security
- Image and Object Detection Techniques
- Micro and Nano Robotics
- Color perception and design
- Infrastructure Maintenance and Monitoring
China West Normal University
2019-2024
Huazhong University of Science and Technology
2019-2024
Hefei University of Technology
2024
Chengdu University of Traditional Chinese Medicine
2024
Science and Technology Department of Sichuan Province
2024
SEGi University
2023-2024
Beijing Institute of Technology
2024
Wuhan Polytechnic University
2024
Henan Provincial People's Hospital
2024
Zhengzhou People's Hospital
2024
Abstract Facial symmetry is a key component in quantifying the perception of beauty. In this paper, we propose set facial features computed from landmarks which can be extracted at low computational cost. We quantitatively evaluated proposed for predicting perceived attractiveness human portraits on four benchmark datasets (SCUT-FBP, SCUT-FBP5500, FACES and Chicago Face Database). Experimental results showed that performance comparable to those with much denser landmarks. The computation was...
Recently, deep supervised hashing methods have shown state-of-the-art performance by integrating feature learning and hash codes into an end-to-end network to generate high-quality codes. However, it is still a challenge learn discriminative for preserving the label information of images efficiently. To overcome this difficulty, in paper, we propose novel Partial-Softmax Loss based Deep Hashing, called PSLDH, Specifically, PSLDH first trains category code each category, will preserve...
Previous studies have pointed out that visual question answering (VQA) models are prone to relying on language priors for answer predictions. In this context, predictions often depend linguistic shortcuts rather than a comprehensive grasp of multimodal knowledge, which diminishes their generalization ability. paper, we propose novel method, namely, KDAR, leveraging knowledge distillation address the prior-dependency dilemmas within VQA task. Specifically, regularization effect facilitated by...
The segmentation of high-resolution (HR) remote sensing images is very important in modern society, especially the fields industry, agriculture and urban modelling. Through neural network, machine can effectively accurately extract surface feature information. However, using traditional deep learning methods requires plentiful efforts order to find a robust architecture. In this paper, we introduce network architecture search (NAS) method, called NAS-HRIS, which automatically on dataset....
Improving safety management and risk evaluation methods is important for the global mining industry, which backbone of industrial development our society. To prevent any accidental loss or harm to human life property, a assessment method needed perform continuous mines. Based on requirements mine evaluation, this paper proposes Pb-Zn model based fuzzy-grey correlation analysis method. The compared with fuzzy TOPSIS Through experiments, results demonstrate that proposed more sensitive has...
Given social networks over time, how can we measure network activities across different timesteps with a limited number of metrics? We propose two classes dynamic metrics for assessing temporal evolution patterns agents in terms persistency and emergence. For each class metrics, implement it using three aggregation models ranging from the most commonly used Average Aggregation Model to more complex such as Exponential Model. argue that problem measuring be formulated Recency Primacy effect,...
Many techniques have been developed for computer vision in the past years. Features extraction and matching are basis of many high-level applications. In this paper, we propose a multi-level features discontinuous target tracking remote sensing image monitoring. The reference pre-extracted at different levels. first-level used to roughly check candidate targets other levels refined matching. With Gaussian weight function introduced, support is accumulated make final decision. Adaptive...
A microrobot with untethered control in 3D space is a good choice to be applied the fields of biomedicine small and confined workspace. In this paper, an electromagnetic actuation system (EMA) which combined Helmholtz coil Maxwell for 5 DOF locomotion built. The magnetic field analysis proposed EMA was analyzed by finite-element-method (FEM) multi-physics COMSOL software. can produce different characteristics such as controllable uniform gradient field, rotating oscillating three-dimensional...
In this paper, mixed metal hydroxide (MMH) was prepared via MgCl2 and AlCl3 by the co-precipitation method characterized XRD, TGA laser particle size analysis. The inhibitory effect of MMH on swelling clay evaluated linear expansion, mud ball, analysis, X-ray diffraction analysis TGA. expansion experiment showed that with a ratio Mg:Al = 3:1 displayed strong bentonite when 0.3% added to drilling fluid, demonstrating better inhibition than 4.0% KCl. Within 48 h, only few cracks were visible...
A mobile edge computing (MEC) paradgim is evolving as an increasingly popular means for developing and deploying smart-city-oriented applications. MEC servers can receive a great deal of requests from devices users, especially in crowded scenes, e.g., city's central business district school areas. It thus remains challenge appropriate scheduling managing strategies to avoid hotspots, guarantee load-fairness among servers, maintain high resource utilization at the same time. To address this...
Long-wave infrared (LWIR) and visible (VIS) cameras can image information at different dimensions, but the way to calibrate these two types of while registering fusing acquired images is difficult. We propose a calibration plate method for thermal imaging solve three problems: (1) inability existing plates address LWIR VIS simultaneously; (2) severe heat interference in cameras; (3) difficulty finding feature points registration due spectra between imaging. Simulation tests error analysis...